WTH? NOS CC Resistors don't measure anywhere close to stated values!?!

[J M Fahey]

Well, I get the analogy: you keep cats as pets (might include dogs, hamsters and goldfish) and no big deal; but you keep a rattlesnake and sooner or later it will bite you.
Same with keeping CC resistors for a long time.

RG hit the nail in the head and I can imagine a mechanism for resistance going down, WAY down.

As accurately described, composition is made out of carbon granules floating in dried phenolic paint or varnish.
Now carbon resistance is quite low, only way to get high values is to have those "apples floating in syrup" to develop high resistance is to have them **barely** touch each other, each contact point having a tiny surface.
Itīs a miracle it actually works !!!!!!!
I can *imagine* a situation where along many years the binder varnish continues drying or curing and contracts, slightly crushing those "apples" against each other, increasing contact surface and lowering total resistance.
Maybe not a big deal on lower values, which start with much higher contact surface, but a problem with highest ones, which start with being *almost* open.
In that case itīs not a moisture problem at all, water is not shorting them , cooking might improve or further degrade them, dunno.

FWIW I made my own potentiometer tracks (surprise !!!) , my supplier was delayed because of lack of personnel (a strike or something) so I donned my blue overalls and helped him (meaning myself )

The process is quite related to CC resistors: he cut long 1" wide strips of Pertinax (phenolic paper) , and covered them with a thin layer of Philips Netherlands supplied "resistive paint" , basically carbon granules floating in uncured phenolic varnish.
Then he cooked them in an oven, and I remember cooking length and temperature was the final fine tuning resistance adjustment, he got within 20% and usually within 10% which for a pot track is excellent.

Then he punched out the omega shaped pot tracks, splashed silver paint on ends to improve contact quality and riveted terminals, then assembled them with sliding contacts inside pot cases .... a fascinating process.

I am certain an insulating tube filled with said conductive paint and cooked would make a fine CC resistor.

Also, silkscreened tracks made out of such paint would make good resistors in any value, I bet thatīs whatīs used inside Custom/Hybrid ICs, such as those STK or Sanken power modules.

It would also be great to make hard to copy pedals or preamps: no regular resistors to read codes from, nor can they be removed for mesuring without being destroyed.
Way better than goop

[Leo_Gnardo]

Also, silkscreened tracks made out of such paint would make good resistors in any value, I bet thatīs whatīs used inside Custom/Hybrid ICs, such as those STK or Sanken power modules.

Late 80's early 90's there was a popular very cheap mix console that did just that: print carbon tracks on the PC board for faders. Made by ADA I think, can't remember the product number.

[g1]

That was an Alesis debacle (1622). Lifespan of the mixer was limited to how ever many cycles those "baked in" carbon tracks could withstand.
Replaceable wearing parts, who needs that?

[Steve A.]

I don't think they were rejects -- I don't think anything would have been kept if it were known to be bad.

Certainly not if they were known to be bad. I was thinking of a bag or box of resistors of assorted values all mixed together of questionable origin which is exactly what I have and exactly what I pulled the resistors from at Mike Quinn's electronics surplus store. (Bill Godbout had worked for Mike in the 70s before going out on his own... I remember ordering a whole sh*tload of RC4739 and RC4136 op-amps from him for building the Craig Anderton kits.)

http://www.imsai.net/history/quinn/quinn-1.htm

https://groups.google.com/forum/m/#!...rs/u4fmiciMr2o

BTW a few years ago I bought two assortments of 1/4W and 1/2W CF resistors from Velleman at the local Frys store to work on SS amps and pedals.*** I don't think I found one resistor within 3% of its supposed value... most were closer to 5% too high or low. My theory was that Velleman was buying batches of resistors which were rejected by QC...

Replaceable wearing parts, who needs that?

Kind sir, I must protest... I resemble that remark!

Steve Ahola

*** For tube amps I have a very expensive assortment of metal film resistors from TRW et al made up over the years with those skinny cards packaged with 2 to 6 pieces for 89 cents... ouch!

[Leo_Gnardo]

That was an Alesis debacle (1622). Lifespan of the mixer was limited to how ever many cycles those "baked in" carbon tracks could withstand.
Replaceable wearing parts, who needs that?

That's the one! I quickly understood it to be "disposable equipment." I'll join you in a .

[The Dude]

Fingers touching the probes maybe? I have to wonder. Don't forget the parallel resistance of "you" when measuring on higher resistance scales.

[J M Fahey]

Fingers touching the probes maybe? I have to wonder. Don't forget the parallel resistance of "you" when measuring on higher resistance scales.

What do you mean saying: "You short everything you touch !!!!!! eh???"

[The Dude]

Well, I didn't say "short", but the human body does have enough resistance to cause inaccurate readings on higher resistance scales. You don't want to be holding both probes with your fingers while trying to measure a 1M resistor if you're concerned about accuracy.

[R.G.]

BTW a few years ago I bought two assortments of 1/4W and 1/2W CF resistors from Velleman at the local Frys store to work on SS amps and pedals.*** I don't think I found one resistor within 3% of its supposed value... most were closer to 5% too high or low. My theory was that Velleman was buying batches of resistors which were rejected by QC...

A lot depends on whether they were sold as 1% or 2% resistors, or were supposed to be 5%.

Back when the quasi-standard for resistors was 20% and 10% was the premium stuff, manufacturers would try mightily to get a batch of a million resistors to come out with an average value dead on the nose of the nominal resistance. So if you measured a million and plotted their values, you'd get a bell shaped distribution with a peak very close to the nominal. The problem was the width of the bell curve. They could not make all the million come in within even 20%, as the process variations were too wide.

So they tested them all. The ones with values outside the 20% were either tossed into the nominal bucket that they were nearest to (i.e. a 1k at -20% makes a pretty good "810 ohm" resistor - you just have to redo the painted rings) or discarded if no other use could be found for them.

But even then, 20% components were regarded among engineers as the stuff you scrape off your shoes with disgust after a walk in a public park. So having tested all million, the makers noted that quite a few of them came in within +/- 10%, and could be sold at a premium price. Hah! The guy who thought that up got a big attaboy from accounting. And it wasn't more than a few nanoseconds before they realized that the 5% ones could be sold for even more, and the 1% ones for more than that.

This led to the middle being tested and re-marked out of the distribution. If you bought 10% resistors, you could be sure that you'd get resistors that were no better than 5%, and that what you actually got was one lump of parts that were between -10% and -5%, and another lump of them at +5% to +10%; no nominal ones at all. Well, until they drifted, but that's another story.

This practice faded out when manufactures got better handles on their process variances and much narrower distributions. Then there was that laser-trimming stuff and suddenly 1% resistors got almost as cheap as 5% resistors.

But your story sounds eerily like the process of testing out the middle of the distribution. So I don't know whether Velleman was buying rejected batches or batches that were lovingly put in the low-dollar bin by truly sharp, on-the-ball QC departments. Velleman may have eagerly sought out the rest of the distribution.

[Enzo]

Oh that horrible Alesis board, I had forgotten it. Now I have a bad taste in my mouth again.

I first saw one, a buddy brought in two of them wanting his controls cleaned. My first clue I was in hell was when I pulled the board off the panel, all the knobs and wiper assemblies stayed on the panel. Just a circuit board with copper traces and black resistor striped painted on. No way to power the board off the panel.

And replace the pots? Impossible to do. Sorry Billy.

[J M Fahey]

Iīve seen "halfway" products: a strip of pertinax with the track printed on one side ... or both for stereo, a sliding contact on a sliding "ski" , simple PCB terminals riveted at each end ... and nothing else, definitely NO case (which is more expensive than the actual working guts):
got dirty easily, but at least could be replaced as a whole assembly.

VERY cheap to manufacture; found them on lower end Car stuff and cheap "mini component" home audio.

[Steve A.]

A lot depends on whether they were sold as 1% or 2% resistors, or were supposed to be 5%.

I just checked and they were sold as 5%. I had an older collection of 1/4W and 1/2W CF resistors going back to the 80's bought in big assortments from Radio Shack and I could usually find some resistors that were spot on, something which didn't happen much with the Velleman resistor kits.

So are the tolerance bands printed before QC or after QC? (You did mention that at one time they would reprint color codes... doesn't sound very cost effective today.)

But your story sounds eerily like the process of testing out the middle of the distribution. So I don't know whether Velleman was buying rejected batches or batches that were lovingly put in the low-dollar bin by truly sharp, on-the-ball QC departments. Velleman may have eagerly sought out the rest of the distribution.

I assume that they bought the cheapest resistors that they could find.

Moral of the story is to get 1% MF resistor kits from Joe Knows Electronics... roughly the same price as the 5% CF kits from Velleman and they come in labelled ziplock bags which fit snugly in a nice cardboard box. $11.99 for 10 each of 86 values. BTW is there any reason to use CF resistors in FX pedals when MF resistors are essentially the same price?

https://www.amazon.com/Joe-Knows-Ele...dp/B003UC4FSS/

For $12.66 you can get 10 each of 164 values if you need the "in-betweeners" as well. Not as well organized but you would have a backup supply if you use up all 10 of a particular value in the JKE kit...

https://www.amazon.com/Top-cofrLD-re...dp/B00WE1FQ8Y/

Steve Ahola

[R.G.]

I just checked and they were sold as 5%. I had an older collection of 1/4W and 1/2W CF resistors going back to the 80's bought in big assortments from Radio Shack and I could usually find some resistors that were spot on, something which didn't happen much with the Velleman resistor kits.

Everything is a variable. Manufacturers did quit testing out the middle of the distribution when they could get tight enough control of the variances so that most of every run wound up inside the tolerance range. I think that "test out the middle" died out mostly in the late 60s as an industry practice. When they stopped it, you started getting resistors that were very close. In fact, my recent experience with carbon film is that "5%" resistors are about +/-2%.

Velleman probably did actively buy seconds. That could be a reasonable, if disappointing, practice. There's about two pages of typing on that topic that I'll spare you.

So are the tolerance bands printed before QC or after QC? (You did mention that at one time they would reprint color codes... doesn't sound very cost effective today.)

Recent (last 30-40 years) practice is to fully automate, make them all identical, then automatically test and trash the defects, of which there aren't many. Color bands are all fully automated too, and that happens any time in the manufacturing flow that they want after the outside passivation/pretty cover layer is dry. Back before the 1960s, having a resistor that was within 20% was good if you could get it, and it was only in the 60s that 10% became common.

Quantities were low and labor was cheap (by today's standards) so additional attention to each part was possible and maybe even practical, as prices were high (again by today's standards). So remarking was possible. Later, 70s through 90s probably, hard-tool automation and rising labor rates made individual attention to each part un-economic, so re-marking was probably abandoned entirely. The rise of cheap computers in the 80s and 90s meant that automated test and handling was possible for individual parts, but it's likely that improved manufacturing and tight process controls got the distribution widths down to where out-of-tolerance parts could simply be trashed; re-marking became unnecessary.

I remember reading a magazine article from the 50s for how to make your own high-precision CC resistors. You took something close, but lower in value than the desired resistance and put an ohmmeter on it that could measure the precision you wanted (no small task with analog needle-meters, BTW) and filed a notch with a triangular file till the resistance came up to your desired value. Then you passivated the filing with paint of some kind.

Moral of the story is to get 1% MF resistor kits from Joe Knows Electronics... roughly the same price as the 5% CF kits from Velleman and they come in labelled ziplock bags which fit snugly in a nice cardboard box. $11.99 for 10 each of 86 values.

A lot depends on how many resistors you use. I was doing prototypes for work for a while there that might need 10-20 each of some values. So I got to buying 5% CF in the first price break per value. That was usually 100 or 200 each, and usually $0.01 to $0.02 per resistor. It's expensive at first, but after a few projects, you mostly don't need to buy resistors at all since you have the other 80 from the last buy. You wind up buying only when the envelope is empty. I keep my resistor stock in coin envelopes in one plastic shoebox. A hundred 1/4W resistors fit in an envelope that is about the width of the box, and the box will just hold the common 5% values between 10 and 2.2M. I mark the envelopes on the corner, so picking out a resistor is like digging a folder out of a file drawer. Probably doesn't work all that great for 1/2W and bigger.

BTW is there any reason to use CF resistors in FX pedals when MF resistors are essentially the same price?

Availability, maybe. Metal film drifts less and has lower noise than CF, but there are places that stock more CF values or have better price breaks to get 50-200 for a few cents each.

[J M Fahey]

I clearly remember than when I started in the late 60`s CF was the new kid in the block and usually came in 10% tolerance, 5% being premium; CC was still available in dealerīs shelves but when one value was finished, restocking meant CF ; we didnīt consider CC as having Mojo of any kind, quite the contrary, itīs what an old semi retired BW tube TV repair guy would have in his parts bin ... bought years before.
Together with his outdated RCA or EICO VTVM or the cherished Hansen needle multimeter, which was a favorite just because it came Factory equipped with a sausage sized HV probe able to measure the CRT 17kV or so.

1% and MF?
I donīt use them even today, just buy the 10000 strips/boxes of 1/4W CF resistors for peanuts in 20% increments from 1 ohm to 3M3 and use that.

HATE going downtown (although itīs 30 minutes away, coming back is another 30 minutes and typically waste 1 hour buying in a couple shops) so I rather work from own stock "my own little Radio Shack " and restock when and if I have to go downtown for something else.

Time spent buying is time wasted, even worse for small inexpensive parts.

[GainFreak]

And what about current production CCs? Are the problems you all mentioned now solved or they will show again over time?